Preoperative sleep disturbance and Postoperative delirium in older Adults: A Mediation Analysis via Insulin-Like Growth Factor-1

Background: Postoperative delirium (POD) is a clinically important complication in older surgical patients. Preoperative sleep disturbance is common and potentially modifiable, yet the biological pathways linking sleep quality to POD remain incompletely understood. We examined the association between preoperative sleep quality and POD, tested whether insulin-like growth factor-1 (IGF-1) partially mediates this relationship, and evaluated the incremental discriminatory value of adding IGF-1 to a clinical prediction model. Methods: This single-center prospective study screened 260 patients aged ≥60 years undergoing elective surgery under general anesthesia at the Naval Medical Center (May–July 2025); 254 were included in the final analysis. Sleep quality in the preceding month was assessed by Pittsburgh Sleep Quality Index (PSQI) and fasting serum IGF-1 was measured. POD was screened daily on postoperative days 1–3 using the Confusion Assessment Method (CAM). Multivariable logistic regression identified independent POD factors; mediation analysis quantified IGF-1 mediation of the PSQI–POD association. Models with vs without IGF-1 were compared using receiver operating characteristic (ROC) curves; areas under the curve (AUCs) were compared with DeLong’s test. Results: Among 254 patients, POD occurred in 45/254 (17.7%). In multivariable logistic regression, higher PSQI scores (OR 1.49, 95% CI 1.26–1.77; p<0.001), older age (OR 1.13, 95% CI 1.06–1.21; p<0.001), and higher C-reactive protein (CRP) levels (OR 1.28, 95% CI 1.14–1.45; p<0.001) were independently associated with increased POD risk, whereas higher IGF-1 levels were independently protective (OR 0.98, 95% CI 0.96–0.99; p=0.038). Mediation analysis indicated a significant partial mediation by IGF-1, accounting for 25% of the total effect (95% CI 0.09–0.74; p=0.002). In ROC analyses, the model without IGF-1 achieved an AUC of 0.82 (95% CI 0.76–0.88), and the model including IGF-1 achieved an AUC of 0.83 (95% CI 0.77–0.89); the difference was not significant (DeLong Z=−1.52, p=0.129). Conclusions : Poor preoperative sleep quality was independently associated with POD in older patients, with IGF-1 partially mediating this relationship. Adding IGF-1 provided limited incremental discrimination. Trial registration: Chinese Clinical Trial Registry (ChiCTR), ChiCTR2500109251 . Registered on 16 September 2025. Retrospectively registered.